18 - Prioritization of Effective Factors on Zataria multiflora Habitat Suitability and its Spatial Modeling

Abstract

Habitat suitability assessment for rare plants is an important component of conservation and development planning. The thyme-like plant Zataria multiflora is a rare species the population of which has been decreasing due to habitat fragmentation and loss. Therefore, the aim of this study was spatial modeling and considering effective factors on Z. multiflora habitat suitability in Fars Province, Iran. For this purpose, firstly, the location of Z. multiflora was detected using extensive field surveys and classified into two datasets, including training (70%) and validation (30%). Then, 13 influence factors, including elevation, slope degree, aspect, soil physical characteristics (sand, silt, and clay percentage), soil chemical properties (EC) and pH), mean annual rainfall, mean annual temperature, distance to roads, distance to rivers, and plan curvature were investigated for modeling aims. In this study, a data-mining technique, namely support vector machine (SVM), was applied in R statistical software to produce a habitat suitability map for Z. multiflora. Finally, to evaluate the accuracy of the developed model, a receiver operating characteristic curve and AUC index were implemented. According to the results, SVM had excellent performance, with an AUC value of 96.5%. Also, the results of variable importance by Boruta algorithm showed that temperature, silt percentage, aspect, distance to road, and EC were the most important factors on habitat suitability of the described species, respectively. On the other hand, pH, sand percentage, distance to river, and plan curvature were the least significant factors, respectively. This model can be used as an effective tool for conservation planning, monitoring, and management of rare plant habitats in the study area.